1 of 47

Qualitative–Quantitative Reasoning�thinking informally about formal things

Alan Dix

http://alandix.com/papers/ICTCS-QQ-2021/

@alanjohndix @CompFoundry

2 of 47

Tiree

Tiree Tech Wave�https://tireetechwave.org/

3 of 47

the foundry

building

mission

community

4 of 47

today I am not talking about …

  • physicality and product design
  • intelligent internet interfaces
  • long tail of small data
  • deep digitality
  • digital thinking
  • now
  • digital light
  • digital humanities
  • virtual crackers and slow time
  • modeling dreams, regret and the emergence of self

5 of 47

Number theory�and things …

6 of 47

7 of 47

8 of 47

9 of 47

arithmetic

mathematics

?

10 of 47

the qualitative understandingof quantitative phenomena

11 of 47

Formal models of �interactive systems

12 of 47

Oxymoron?

𝛌

?

=

https://en.wikipedia.org/wiki/File:The_Adicts_2011_SO36_03.jpg

13 of 47

PIE model – abstract properties

predict

e.g. predictability:

14 of 47

undo

15 of 47

more complex undo – the cube

16 of 47

PIE model – abstract properties

predict

e.g. predictability:

17 of 47

Insight!

Formal model often not complete or accurate

… but still gives insight

predict

18 of 47

Fully Charged

19 of 47

Electrostatically-charged crop sprays

20 of 47

Above canopy model

ground at�0 V

open boundary condition complex

sprayer at �high voltage

+ 1000s V

charged�spray drops

model path, speed and space voltage�due to drop cloud

?

21 of 47

Within Canopy Model

crop modelled as�straight sides

ground at �bottom

+V – charged spray enters crop canopy from above;�speed and charge determined by above canopy model.

22 of 47

Within Canopy Model

crop modelled as�straight sides

ground at �bottom

+V – charged spray enters crop canopy from above;�speed and charge determined by above canopy model.

23 of 47

Misses crop (Class I)

24 of 47

Good spread (Class II)

25 of 47

All at the top (Class III)

26 of 47

Connect and re-parameterise

+ 1000s V

above crop model

within�crop model

dimesionless�parameters

27 of 47

Design/engineering interventions

X

28 of 47

Order of magnitude�day to day

Ocean and drops

Covid and climate change

29 of 47

an ocean is made of drops

Covid – masks, etc.

    • action action minimal impact
    • together reduce growth

climate & pollution

    • each car ride, plastic cup, …�minimal difference
    • together, each year … 1.2 trillion tonnes of ice melt�8 million tonnes of ocean plastic

30 of 47

Order of magnitude�formal

Algorithmic complexity

Linear programming and sorting

What is computation?

31 of 47

what is computation?

information theory� entropy independent of mechanism

algorithmic complexity � measures are system dependent

sorting? – n log n lower bound� information proof – works for oracles

Galois theory feels connected?

32 of 47

standard Galois�field extension

(√2)

computational �Galois sets

0(√2)

(√2)

2(√2)

1(√2)

3(√2)

=

=

reals

rationals

rationals �extended �with solution �of x2=2

three √2 allowed

plain rationals

two √2 allowed

one √2 allowed

any number of �√2 allowed

33 of 47

Numerical models

As well as crop spraying …�Covid serial interval

34 of 47

Viral metrics

R – how many?

Serial interval – how long?

time

35 of 47

Viral metrics

R – how many?

Serial interval – how long?

not plain disease metrics …

  • about disease in a particular setting:� housing, social culture, precautions
  • interlinked

time

36 of 47

Serial interval – 3 measures

Forward (standard)

    • How long to next infection

Backward

    • How long since infected

Effective

    • During exponential growth

R

t

SI

37 of 47

Infection time not uniform

⇒ during growth (R > 1)

backward SI < effective SI < forward SI

38 of 47

Monotonic reasoning

Shops to politics

39 of 47

Monotonic arguments

Change in shop

A – b < A

2D – Poincare property

40 of 47

Factory Automation

41 of 47

more�automation

less�labour needed

higher�productivity

higher�prosperity

less�employment

lower�wages

? people�better off

42 of 47

more�automation

decline of�other companies

more�competative

growth�of company

less�jobs elsewhere

more�jobs there

? overall�employment

43 of 47

Formalising QQR

Fuzzy logic, Bayesian methods

Allen’s interval algebra

44 of 47

Allen's interval algebra

A

B

A

B

precedes

meets

A

B

A

B

starts

finishes

A

B

A

B

contains

overlaps

A

B

equal

45 of 47

Call to arms

46 of 47

QQR critical for understanding …

climate change

Brexit

in short ….� being a C21 citizen

covid

47 of 47

Alan Dix

http://alandix.com/QQR/

@alanjohndix @CompFoundry

Qualitative–Quantitative Reasoning�thinking informally �about formal things